From the course: Algorithmic Trading and Stocks Essential Training

Unlock the full course today

Join today to access over 24,000 courses taught by industry experts.

Evaluating models

Evaluating models

From the course: Algorithmic Trading and Stocks Essential Training

Start my 1-month free trial Buy for my team

Evaluating models

“

- [Instructor] It's not enough to just be able to come up with an effective algorithmic trading model, we need to be able to adjust that model over time as circumstances dictate. I'm in the 0308 Begin Excel workbook. So what we've got here is a model for predicting light vehicle sales, and we'd predicted these light vehicle sales based upon different factors or variables, including gas prices, Moody's BAA bond yield averages, initial jobless claims, crude oil prices and industrial production. And we spit out the regression that you see here, which was reasonably accurate, .73 R squared, adjusted R squared, pretty good. And that led us to our prediction for what vehicle sales would be over the next six months based on different assumptions about those economic variables, and we could compare that to what it had been on average over time, but what if we needed to improve this model? What if we decided that the model…

Contents